A Retrospective Cohort Study Comparing Utilization and Costs of Biologic Therapies and JAK Inhibitor Therapy Across Four Common Inflammatory Indications in Adult US Managed Care Patients

Many inflammatory conditions are severe, chronic, and disabling diseases that can manifest in joints and skin. Rheumatoid arthritis (RA), which affects approximately 1.5 million adults in the US and an estimated 0.5–1% of people in developed nations, primarily affects the synovial membrane in the joints [1]. Psoriasis (PsO) is a skin disease occurring in approximately 4.5 million adults in the US, with an estimated prevalence of 1.4–2.5% in children, adolescents, and adults in the US [2]. PsO is frequently accompanied by comorbid psoriatic arthritis (PsA), a disfiguring form of arthritis with an estimated prevalence of 11% in patients with PsO in the US [3]. Ankylosing spondylitis (AS), with prevalence estimated to be 30–900 per 100,000 people in the US, is a type of arthritis that affects the spine [1]. In addition to symptoms associated with these diseases (pain and swelling of joints and plaque formation on skin), patients suffer impaired health-related quality of life [4, 5].

Biologic therapies that target specific components of the immune system are important for treatment of autoimmune diseases. These include agents that target tumor necrosis factor (adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab), interleukin (IL)-6 receptor (tocilizumab), CD20 (rituximab), IL-12 and IL-23 (ustekinumab), and T cell stimulators (abatacept). A recently approved non-biologic therapy that targets JAK (tofacitinib) is also used to treat RA; this medication has a similar chemical profile, price, and target population in the product label relative to the biologics, and is therefore included in this analysis. These drugs differ in mode of administration, dosing level, dosing schedule, and whether they are approved for first-line use after non-biologic disease-modifying antirheumatic drug (DMARD) failure. Adalimumab, certolizumab pegol, etanercept, golimumab, and ustekinumab are administered as subcutaneous (SC) injections; infliximab and rituximab are administered as intravenous (IV) infusions; and tofacitinib is administered orally (PO). Abatacept, tocilizumab, and golimumab are available for both SC and IV administration.

The cost of biologic therapies is high compared to non-biologic DMARDs; evidence-based assessment of comparative costs between biologics is critical to ensure rational healthcare resource allocation. Data comparing costs across indications from the payer’s perspective and an understanding of biologic dosing patterns across different populations and payers are critical for formulary decision-makers to develop evidence-based formularies. Although prior studies evaluated the comparative costs of biologics across multi-payer claims databases, the data used in those studies are now relatively outdated and such analyses of multi-payer data may be less applicable to specific payers. We now report the results from a biologic utilization model using more recent utilization data from claims for beneficiaries on a biologic enrolled in a health plan of a large national commercial health insurer.

This was a retrospective analysis of total annual cost of biologic treatment and biologic utilization among patients diagnosed with at least one of four autoimmune conditions: RA, PsO, PsA, or AS. The most commonly used medications were etanercept, adalimumab, and infliximab, which is consistent with previous findings [6‐8]. Notably, the most common medications had been approved longer than the other biologics included in the study. While most of the patients were considered to be continuing on existing biologic therapies, a few medication cohorts (rituximab, certolizumab, and tofacitinib) were more often new users.

Etanercept was the most commonly used medication in this study and in a prior study [9]. Similar to the prior study, newer biologics had equivalent or lower costs than etanercept for RA, PsA, and AS. The current study also confirms the observation that for biologics that are indicated for all four inflammatory diseases (etanercept, adalimumab, and infliximab), etanercept had the lowest annual acquisition and administration costs [9].

This study has inherent limitations associated with observational studies based on administrative claims. Specifically, while claims data indicate whether a prescription was filled, they do not indicate if the medication was used as prescribed or at all. This study included only patients who were enrolled in a commercial health plan; results of this analysis may, therefore, not be generalizable to outcomes in patient populations with government-sponsored health insurance (e.g., Medicare, Veterans Health Administration coverage, state Medicaid coverage) or uninsured populations. Results are also not generalizable to patients outside the US, as treatment guidelines and costs vary across countries. Additionally, claims data lack clinical data that may impact the interpretation of the results. Specifically, the duration of clinical benefit is assumed to be constant for each of the infused medications, which could lead to overestimates of persistence for certain medications. For example, the estimated duration of clinical benefit for the last infusion of infliximab may be too long, since the label-recommended dosing frequency could vary from every 8 weeks to every 4 weeks, which may impart bias by overestimating the clinical benefit for the last infusion. Similarly, the total duration of treatment for continuing users is unknown, leading to potential bias if it differs by index medication. Important clinical information such as reasons for treatment modifications (switch, restart, discontinuation), and clinical response to the biologic is unavailable in a claims database. Claims data also do not contain biometric information, such as patient weight and thus weight was not incorporated into dosage calculations for infused biologics. ICD-9-CM codes are proxies for actual physician diagnoses and can be affected by miscoding, overcoding, and undercoding. All variables may be subjected, to some extent, to data entry errors or errors due to rounding (for numerical variables). These errors, however, are unlikely to differ across groups or outcomes. Finally, this study calculated healthcare cost based on the observed dose and the associated WAC price, which does not take into account negotiations between health plans, manufacturer’s rebates, or other price modifications that could impact the actual health plan and patient paid amounts.

The findings from this study indicate that there is considerable variability in the 1-year cost of biologic treatment between medications as well as between indications and in new or continuing patients. Etanercept was shown to have lower costs relative to adalimumab and infliximab across the indications evaluated here except among patients with PsO only. Of the medications included in this analysis, ustekinumab was the most expensive therapy for PsO. Newer agents may be associated with lower treatment cost; however, future research is needed to see if these estimates persist with expanded use of the more recently approved biologic.